RTG: Modeling and Uncertainty Quantification for Life Sciences

North Carolina State University, Raleigh NC

Investigators

Abstract

Mathematical models of physical and biological systems are foundational in life science. In order for these models to provide reliable information for decision-makers, they must be rigorously analyzed and calibrated using advanced statistical and mathematical methods. The RTG project, Modeling and Uncertainty Quantification for Life Sciences (UQ4Life), will train a diverse group of students and postdoctoral fellows with the technical skills needed to advance mathematical and statistical research in UQ and the soft skills required to lead interdisciplinary teams. All research conducted by trainees will address important scientific problems and be conducted in collaboration with domain experts, ensuring that new knowledge will be disseminated beyond the mathematical and statistical communities. To extend the impact beyond NC State, the project includes a Distinguished Seminar Series and UQ Hackathon that are shaped and conducted by UQ4Life trainees and designed to engage with leading UQ researchers. The ambitious program is organized around three specific aims: (1) train data scientists with core proficiency in modeling and UQ methodology, (2) create an interdisciplinary culture to solve important life-science problems, and (3) strengthen and diversify the STEM workforce. To achieve these aims, the PI team have assembled a vertically-integrated team of researcher mentors from the Statistics and Mathematics Departments at NC State. Trainees will be given professional development mentoring, networking opportunities, and training and experience in communication and leadership. The interdisciplinary training provided by UQ4Life will result in significant advances in both methodology and application of modeling and UQ. Trainees will develop new methods for modeling life-science systems, Bayesian methods for model calibration and processing large datasets, and equation learning methods that use machine learning concepts to estimate dynamic non-linear relationships between state variables. These new methods will be motivated by important applications spanning the areas of climate, physiology, ecology, and disease modeling. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.